Report: The world's first AI regulation, the EU's Artificial Intelligence Act（AI Act）》At local timetwo thousand and twenty-fouryeareightmonthoneIt officially takes effect on [date]. This is the world's first comprehensive regulation on artificial intelligence, which establishes unified rules for the development, market deployment, and use of AI systems, aiming to ensure the safety and credibility of AI technology, protect basic rights, and establish a fair and unified AI market within the European Union.

In 2021, the European Commission first proposed a negotiation authorization draft for the Artificial Intelligence Bill proposal. After multiple rounds of discussions and revisions by the three core EU institutions, the European Parliament, the European Council, and the European Commission. 2023twelvemontheightThe EU and Japan have reached an agreement on the Artificial Intelligence Act; And in 2024twomonthtwoObtaining EU membership on a daily basistwenty-sevenUnanimous support from national representatives;fivemonthtwenty-oneThe European Council officially approved the bill until AugustoneFrom now on, the bill will officially come into effect.

two thousand and twenty-threeyeartwelvemontheightJapan and the European Union reach an agreement on the Artificial Intelligence Act

The Artificial Intelligence Act mainly defines and classifies artificial intelligence, clarifies regulatory requirements and prohibited behaviors, also requires transparency of artificial intelligence, and stipulates corresponding punishment measures.

The bill defines the scope of artificial intelligence systems and classifies them into four categories based on risk levels: prohibited risk, high-risk, limited risk, and low-risk. Strict regulatory requirements will be implemented for high-risk artificial intelligence systems, including risk assessment and human supervision. For generative artificial intelligence tools (such asChatGPT）Implement control measures to ensure transparency and accuracy in their use. Explicitly prohibit certain AI practices that violate EU values and are particularly harmful, such as using AI for "social scoring" and manipulating human behavior. Require artificial intelligence systems to have sufficient transparency to explain their decision-making process and results. Introduce transparency obligations for general artificial intelligence models, including compliance with EU copyright laws and publishing detailed summaries for training content. Finally, it was stipulated to impose high fines on enterprises that violate regulations, with the amount of fines depending on the nature of the illegal behavior involvedAIDetermine the level of risk in the system and the global revenue of the enterprise.

The scope of application of the Artificial Intelligence Act is broad, covering any product or service that uses artificial intelligence systems. It not only applies to enterprises within the European Union, but also has extraterritorial effects, binding on enterprises outside the EU that provide services or products to the EU market. Specifically, the bill applies to all artificial intelligence system providers, users, importers, distributors, and product manufacturers with a connection point to the EU market.

Experts say that the EU's Artificial Intelligence Act is a landmark legislative measure that not only reflects the EU's ambition to regulate the field of artificial intelligence, but also reflects a profound understanding of the challenges and opportunities brought by the rapid development of artificial intelligence technology worldwide. As the world's first comprehensive regulation on artificial intelligence, it will have a profound impact on the regulation of artificial intelligence worldwide. The laws and regulations of the European Union have always been considered to have“Brussels Effect”The larger and more affluent a country's consumer market is, the more likely its export enterprises are to comply with EU standards, thereby driving global growthAIStandardization of regulatory standards.

Although the EU Artificial Intelligence Act has come into effect, its specific implementation details and regulatory standards still need to be further improved and refined. With the continuous development of AI technology and the expansion of application scenarios, the European Union and other countries may revise and adjust regulations according to actual situations. In the context of globalization, cooperation and coordination among countries in regulating artificial intelligence have become particularly important. The implementation of the EU's Artificial Intelligence Act has global implicationsAIRegulation provides important references and inspirations, and in the future, countries willAIRegulatory cooperation will also be further strengthened.